Two variables are said to be covariates if they are related to each other and change together. A covariate changes along with the dependent variable in our regression/ANOVA model.
For example, suppose we are studying the effect of the number of exercise hours on weight loss of people. Then the number of exercise hours is a covariate because changes in the exercise hours correlate with changes in weight loss levels.
Weight loss also depends on the initial weight of the person. Hence initial weight is also a covariate.
Types of Covariates:
Covariate variables are generally of two kinds – Independent Variables and Confounding variables.
The independent variables are the variables that are directly under study such as the number of exercise hours in the above example.
The confounding variables are the hidden variables that the researcher may have not considered. An example would be the initial weight of people in the above example. We use the ANCOVA technique to deal with the presence of confounding variables in our study.
Examples of Covariates:
- Quality of soil and growth rate of crops.
- Smoking habits and risk of cancer.
- Diet plan and weight loss.
- Age and shoe size.
Coefficient of Correlation between two Variables:
We can measure the degree to which two variables are covariates by calculating the coefficient of correlation of the two variables.
The coefficient of correlation is a numerical quantity that lies between -1 and 1 and it tells us how much the two variables are related to each other.
If the correlation coefficient lies near +1 then there is a strong positive correlation between the two variables and if it lies near -1 then there is a strong negative correlation between the two variables.